转子分布不平衡和轴承特性参数同时辨识方法

Simultaneous identification method of rotor distributed unbalance and bearing characteristic parameters

摘要:

针对转子模态动平衡的关键参数转子不平衡量和轴承特性参数难以确定的问题,提出一种基于轴承处振动响应的轴承特性参数和转子分布不平衡参数的同时辨识方法。以多项式函数描述转子质量偏心曲线,表征转子连续分布的不平衡质量；以有限元法建立转子/轴承子结构运动微分方程；采用虚功原理将连续分布的广义不平衡力转化为节点处的集中不平衡力并建立其与偏心曲线系数的关系,从而推导出轴承特性参数及质量偏心曲线系数的同时辨识方程,进而实现基于不同转速下轴承处的振动响应辨识出轴承特性参数和分布不平衡参数。以一个转子轴承系统为例进行了偏心曲线阶次对参数辨识的敏感度仿真,结果表明:偏心曲线阶次对参数辨识结果影响较小。以实测轴承处响应辨识得到的参数进行一阶模态动平衡实验,平衡后轴承处振动明显降低,验证了提出的辨识方法及其在模态动平衡中的有效性。

Abstract:

Aiming at the problem that the key parameters of rotor modal dynamic balance, which are rotor imbalance and bearing characteristic parameters, are difficult to determine, a method for simultaneous identification of bearing characteristic parameters and rotor distribution unbalance parameters based on the vibration response at the bearing was proposed. This method used a polynomial function to describe the eccentricity curve of the rotor mass for characterizing the continuously distributed unbalanced mass of the rotor; established the motion differential equations of substructures in the system with the finite element method; converted the continuously distributed generalized unbalanced force into the concentrated generalized unbalanced force of nodes through the principle of virtual work and established the relationship between concentrated unbalanced force and eccentric curve coefficients, thereby deduced the simultaneous identifying equation of bearing characteristic parameters and mass eccentricity curve coefficients, and then realized the identification of bearing characteristic parameters and distribution unbalance parameters based on the vibration response of the bearings at different speeds. Taking a rotor-bearing system as an example, the sensitivity of the eccentric curve order to parameter identification was simulated, the results show that the order of the eccentric curve has little effect on the parameter identification results. The first-order modal dynamic balance experiment was carried out with the parameters identified from the response of the measured bearings and vibration at the bearings significantly reduced after balancing, which verifies the proposed identification method and its effectiveness in model dynamic balance.